gh-stack

Manages stacked PRs with gh-stack — creation, viewing, editing, push, submit, sync, rebase, merge, and checkout. Use when splitting work into reviewable branches, managing dependent PRs, whenever the user mentions a stack, branch layers, or gh stack, or whenever a stack is already checked out.

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

The canonical home for this skill is gh-stack in flightctl/ai-workflows

Quality

Content

80%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a lean, well-structured router that practices progressive disclosure exactly as intended: an overview plus trigger-conditioned, one-level-deep references with no token waste. Its weaknesses are inherited from its minimalism — no inline quick-start command and no visible validation checkpoint for the destructive merge/rebase/push workflows it routes to, which caps workflow clarity.

Suggestions

Add one validation checkpoint to the body for destructive workflows, e.g. "Before merge/rebase, confirm the stack is in sync (see commands.md preconditions) and re-verify after push" — this lifts workflow clarity past the rubric's batch-operation cap.

Inline a single quick-start command (e.g. the `gh stack create` or `gh stack submit` invocation) so the most common case is executable without opening a reference.

State what success looks like after a push or submit (e.g. expected PR links or updated stack status) so failures are detectable at the body level.

Dimension	Reasoning	Score
Conciseness	The ~15-line body contains only routing: the single context sentence ("`gh stack` is a GitHub CLI extension for stacked branches and pull requests") establishes non-obvious tool identity, and every other line directs a file read. No explanation of concepts Claude already knows, no padding.	5 / 5
Actionability	Concrete, conditional routing ("Read [references/usage.md]... for setup, the non-interactive command table, core workflow loops, and exit codes"; "Open the reference whose trigger matches the task") tells Claude exactly which file to open when, but no command or example appears inline — the executable content is entirely delegated to the references, leaving it just short of the 5 anchor's "specific examples cover the common cases".	4 / 5
Workflow Clarity	The body sequences reads clearly (usage.md first, then trigger-matched reference), but the skill manages push, rebase, and merge across stacked PRs — batch/destructive operations — and the body itself shows no validation or verification checkpoint. Per the rubric cap, a destructive/batch skill without validation cannot score above 3.	3 / 5
Progressive Disclosure	A clear overview with four well-signaled, one-level-deep references (all verified to exist: usage.md, stack-design.md, commands.md, troubleshooting.md), each paired with an explicit trigger condition ("before creating a stack", "on unexpected failures", "on conflicts, divergence, restructuring"). Easy navigation, no nesting, nothing inlined that belongs in a separate file.	5 / 5
	Total	17 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is exemplary: it names the tool and domain, enumerates the complete set of concrete operations, and pairs them with an explicit, trigger-rich "Use when" clause in third person. It matches the rubric's good examples in structure and quality with no padding or over-claims.

Dimension	Reasoning	Score
Specificity	"creation, viewing, editing, push, submit, sync, rebase, merge, and checkout" lists nine concrete actions covering the tool's full command surface — comprehensive coverage per the 5 anchor, not 4, since there are no minor gaps.	5 / 5
Completeness	The what ("Manages stacked PRs with gh-stack — creation, viewing, editing, push, submit, sync, rebase, merge, and checkout") and the when ("Use when splitting work into reviewable branches... whenever a stack is already checked out") are both explicit and concrete, mirroring the 5 anchor.	5 / 5
Trigger Term Quality	"whenever the user mentions a stack, branch layers, or gh stack" plus "splitting work into reviewable branches" and "managing dependent PRs" covers the natural phrasings and their synonyms users would actually say.	5 / 5
Distinctiveness Conflict Risk	A clear niche — stacked-PR management via a named tool — with distinctive triggers ("stack", "branch layers", "gh stack") that would not fire for generic branch or git skills.	5 / 5
	Total	20 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: amir-yogev-gh/ai-workflows
Path: skills/gh-stack/SKILL.md
Commit: 2bd6607

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.